A French research institute, CEPII, produced a report on the servitization status of French manufacturing companies last year and since we haven’t said much about France so far in this blog, we decided to summarize some of the findings here. We’ll also archive the paper in our research section.

First of all it’s probably best to ignore the paper’s pessimistic and statist tone. At least implicitly the authors associate servitization with de-industrialization and societal decay of sorts, lament the decline of manufacturing, regard services as a kind of poor cousin and express doubts on whether services can deliver sustained growth. I don’t expect this to be the prevailing opinion in France, however a comparative analysis of what people think about servitization in various countries would be interesting -and in fact could be a proxy servitization success indicator. It is one more example of why not all markets are the same and why in different markets different strategies are required.

But to concentrate on the findings which are based on data from 68,000 manufacturing firms out of a registered 635,000, an impressive sample:

It turns out that in 2007 (latest year for which data are available), services accounted for 11.4% of aggregate sales while 83% of firms sold at least some services, 40% sold more services than goods and 26% did not even produce any goods (though that raises the question of why then they are classified as manufacturers). The average firm-level share of services as a percentage of total sales was a very respectable 35%, which however was very unevenly distributed: In 2/3 of firms services accounted for less than 20% of sales while in 30% of firms for over 80% as can be seen from the reproduced graphic -and this applies to a wide range of manufacturing industries, incl. textiles, machinery, metal fabrication and printed and recorded media (examples under b). In other words there are many companies with very limited servitization, some companies with a lot and almost nothing in between. The authors essentially think this is a quirk in the data and that the companies exhibiting very high levels of servitization are essentially service companies that were once manufacturers, but have failed to de-register as such.


Interesting is also servitization development in the ten years to 2007. The authors define servitization as change in service intensity over time, service intensity being the ratio of service sales to “production sales”, the latter probably meaning “product sales” -as good an operational definition as any. Here the study finds only limited “progress” of less than 2-3% p.a overall, with the change in intensity within companies (“within” column in the table) being much stronger than the change in intensity among companies (“between” column in the tables) – i.e. comparative intensity.

french cos service intesity_2 french cos service intensity_1

Because of this the authors conclude rather interestingly, though not necessarily persuasively, that: “The main driver behind the servitization of the French manufacturing sector is not that highly servitized firms performed better than less servitized ones. It is that each manufacturing firm shifted away from the production of goods to the production of services”.

The closing remarks reflect the view (again implicitly though quite plainly) that servitization is not in fact a forward looking strategy for the future associated with innovation, new business models and capabilities or higher productivity, but essentially a side effect of the decline in manufacturing: Companies are not competitive in making products any more, so they provide services to arrest the decay. This may mirror the state of economic sentiment in France however, more than it does the state of servitization.


A survey by Bosch Rexroth meanwhile paints a pretty dire picture of the state of UK maintenance practices. According to the findings maintenance departments are under-resourced, most companies conduct maintenance in a fairly unsophisticated and unproductive ways (reactive, preventive), use KPIs that are not very… indicative, do not train enough and cannot find sufficient numbers of experienced engineers.

All of this may of course be true, but it is definitely not new. A survey conducted 25 years ago would have produced similar if not the same results.  The problem with this type of surveys which rely on peoples’ answers to questionnaires is that they reflect a particular context in time and the answering person’s relative perception of his/her situation. Maintenance managers will always complain about their budgets and that they don’t have enough well qualified people – simply put resources are scarce. So not much new under the sun as the Assyrians of 5000 years ago remind us:

“Our earth is degenerate in these latter days; there are signs that the world is speedily coming to an end; bribery and corruption are common; children no longer obey their parents; and every man wants to write a book.”
-from a tablet preserved in Istanbul.

More interesting is the fact that there does not seem to be much improvement in predictive maintenance practices (why not?) or the fact that OEMs continue to dominate the maintenance market: The indicated 58% market share is similar to what broader international surveys indicate and that number is something that should be followed closely by observers of service markets: it will show whether third party IIoT based service providers will succeed in taking business away from OEMs.

Germany v US in the Industrial Internet of Things

In the meantime in Germany, Industrie 4.0 is that country’s answer to the Industrial Internet of Things (GE, Cisco, IBM, AT&T, Intel etc – mainly US companies) driven mainly by a fear (quite real) that US software firms (Google, Amazon) will somehow master manufacturing and render German companies, most of which are medium sized niche market champions with high margins, irrelevant. In this dystopian scenario software based service providers practically usurp differentiation of products (differentiation in the service rather than the product) reducing product manufacturers to the second and third division  as makers of commodities with low margins and even lower prestige. We have written a couple of posts about this risk, also in this blog, for example here or here, and we will write about it again. It is worthwhile considering that the German car manufacturing industry provides jobs, directly and indirectly, for more than a million people, while the people depending on it and its global profits for their livelihoods are a sizable fraction of the German population. What happens if, in a servitized world, car-sharing with self-driving vehicles managed by large fleet owners takes hold? How will these global profits fare? It seems to me not very well, given that the value will be in the software and that fleet owners/managers have no particular incentive to buy many expensive cars. Commodities indeed.

Industrie 4.0 was initiated a couple of years ago by the German government and according to a very interesting article in Bloomberg:

“At stake is the health of German manufacturing, which employs 15 million people, about a third of the workforce. By 2020, Industrie 4.0-related projects will account for half of capital investment by German manufacturers, or some $45 billion, according to PricewaterhouseCoopers. Globally, investment in the industrial Internet will top $500 billion a year in 2020, up from $20 billion in 2012, researcher Wikibon estimates. To avoid falling behind, the “Mittelstand must maintain contact with the customer and not lose out” to software companies that might end up with valuable market data, says Volker Treier, deputy head of the German Chambers of Commerce & Industry”.

Bloomberg then highlights the differences between the Industrie 4.0 initiative, mainly a cheer-leading and applied research exercise by the government and other participants in industrial policy and the Industrial Internet Consortium, a private sector initiative coordinating standards and technology trials:

For all of Germany’s concern about maintaining its edge in the new industrial economy, Industrie 4.0 acts primarily as a cheerleader and offers little in the way of financial help. Politicians and labor representatives have a strong hand in setting the group’s agenda, which focuses mostly on sponsoring research at top universities.

While the IIC appears to be out in front, Germany’s approach could lead to greater progress down the line as the adoption of greater connectivity in manufacturing could take decades, says Rainer Glatz, who leads Industrie 4.0 projects at the VDMA German machine makers’ association. “In the U.S. they want to make lots of small steps as quickly as possible,” Glatz says. “In Germany, the effort is far more theoretical: Find the model first and then move toward implementation.”

The latter view seems to me quite complacent by design. In the face of rapidly accelerating technological change how can you expect to “find a model”. And how do you drive business model change towards services if that is not rooted in practice, trial and error and culture change? The next decade will be challenging for many German companies.